
When a system goes down, the business doesn’t just ask when the server will be back up. It asks when it can return to work, and how much of the new information will still be there. Natural disasters, cyberattacks, and failures can disrupt operations and cause data loss. Therefore, data availability and business continuity need clear targets, not just the reassuring promise 'we'll handle it quickly.'
The key principles
- Define RTO and RPO
- Test recovery in practice
- Coordinate people and processes
It’s necessary to define what counts as a return to work: available data set, functioning application, operational system, or the entire organization. The two metrics that help bring order are RTO — recovery time objective, and RPO — recovery point objective. They define how much downtime and data loss the business is willing to accept.
When the targets are clear, it’s easier to prepare, reduce operational impact, and maintain customer trust. But the business changes, and the targets need to change with it. Those responsible for recovery must improve the people, processes, and technology to meet them, and sometimes recover faster and with less data loss.
Let's start with an introduction to RTO and RPO, then move on to four working methods that can help meet them. No test at the end — but there is a recovery check.
Two simple questions: how long, and how much information?
It's easy to explain why recovery objectives are important to a business. The more interesting part is how to translate them into backup and recovery that actually work. Here is the goal of each metric and the way it impacts planning.
RTO: Recovery Time Objective
RTO answers the question: within what time should the data, system, or service be restored? This is the time the business is willing to allocate for recovery after a disruption, before the downtime turns into an unacceptable impact.
To determine RTO, consider how critical the service is and the financial and reputational cost of prolonged downtime. A target that is not met may leave the business without essential operations far longer than planned.
For example, a company in the financial services sector may have an RTO of a few minutes for a central transaction system, and a longer target for a less critical system.
RPO: Recovery Point Objective
RPO answers a different question: how much new information can we afford to lose? This is usually described in terms of time—for example, changes that have occurred since the last proper backup. The goal is to define a loss that the business can absorb without severely impacting continuity.
Turning on the server is an important step, but you also need information that allows work to continue. A system that comes up quickly with missing data can still leave the staff without what they need.
How do you set the targets?
Conduct a Business Impact Analysis (BIA). It examines the importance of activities, the downtime that can be tolerated, and potential data loss, translating these into objectives. It is likely that different business activities will have different recovery objectives.
What you want to recover tomorrow needs to be properly backed up today
The objectives are not just a matter of the restoration action. They affect the way backups need to be prepared in advance. Setting an RTO for a server running an application and a database is not sufficient on its own.
You can write in the plan "two minutes" by the organization's large data repository. For it to actually happen, a backup method and infrastructure capable of restoring it within that time are required. Otherwise, we have a nice goal in the document—and a clock that doesn’t care about it.
The shorter the RTO, the more appropriate and faster the method required for service restoration. If the target is long, server images stored on tape or in the cloud may be suitable. If the target is measured in seconds, an alternate interim environment with continuous replication may also be required.
Similarly, a long RPO can allow less frequent backups. A short RPO may require higher frequency, a different backup method, or a combination with replication. The source gives as an example a transition from file backup to image backup.
Four ways to make objectives practical
Meeting targets requires continuous improvement. As the organization grows, data and infrastructure change. Therefore, it is necessary to examine how to improve performance to meet the targets and even achieve better results.
Beyond just creating a recovery plan, the source suggests four working methods.
Conduct a business impact analysis on a regular basis.
BIA allows understanding the importance of actions in terms of downtime and data loss. As the business changes, the impact of information, applications, and platforms on it also changes.
Periodic review and updates should reflect changes in processes and infrastructure. Use the findings to define the required RTO and RPO while considering budget and technological constraints. Such a review, together with reviewing existing objectives, helps in two ways:
- Backup and recovery teams will have objectives aligned with business needs.
- If it is not possible to meet a new objective due to lack of budget or technology, this can be escalated to management to formulate a solution.
Objectives are validated through recovery exercises.
The source cites a statistic according to which almost 90% of organizations have a recovery plan, but only 50% perform a full DR test every year. It attributes the statistic to Forrester, State of Disaster Recovery Preparedness, 2024. The statistic is presented as the provided document data, not as a new measurement.
Every organization needs to run exercises to know whether it can truly recover. Beyond that, the exercise allows explicit verification of the objectives. Simulate scenarios and check whether the RTO and RPO of each business activity are achievable.
Use exercises to find bottlenecks and gaps in processes and technologies that were assumed to be suitable but do not meet the need. Verifying the objectives helps refine the plan and ensure they are realistic. If you are among the organizations that do not conduct checks, this is another reason to prioritize it.
Improving monitoring and event detection
Recovery time includes more than just copying data from backup. You also need to detect that a disruption occurred, understand what happened, and start taking action. If the fault waits for an hour before someone notices, even very fast restoration begins late.
Early detection allows recovery to start sooner and shortens the time to restore services. Use monitoring tools that provide alerts on anomalies in application, service, and system availability. Integrate them into response procedures to optimize handling.
Ensuring effective communication and training
Even a DR plan with excellent technology depends on people. Every role holder needs to know what to do before an event, during it, and after. Without coordination among participants, it is difficult to meet goals—even if everyone is very busy.
Beyond technology, address three areas:
- Roles and responsibilities: Who coordinates the work, who restores, and who updates the organization.
- Communication plan: How to inform employees about downtime and the situation, and what to do regarding external communication, the press, and social media.
- Exercises and after-action reviews: Conduct periodic training for all participants and table-top exercises that simulate coordination among people, even separate from testing technical restoration.
Clear communication channels and regular training help everyone understand their role and reduce confusion and real-time delays.
Start with improvements that can be measured
RTO and RPO are the foundations of an effective recovery strategy. Improving backup and restoration operations in relation to them can strengthen resilience and reduce potential loss.
The recommendations in the document are intended to help organizations improve recovery performance and their ability to meet targets. Implementing them supports preparedness for unexpected challenges and business continuity in a changing environment.
Choose one critical system, review its objectives, and measure recovery. A gap between the plan and the outcome is useful information: now you know what needs to be improved, instead of discovering it during downtime.


